Compound Q is a purified preparation of a protein found in the root of Trichosanthes kirilowii (the Chinese Cucumber), which has been used traditionally to induce abortion and to treat some forms of cancer in China.

The exact mechanism of action against HIV is uncertain but it is reported to selectively inhibit viral replication and then kill infected cells in vitro. It is this latter effect which makes it different from other therapies. It actually eradicates infected cells, so there are theoretical grounds for hoping that it might clear HIV altogether (i.e. compound Q may be virucidal rather than merely virustatic). No other agents are yet known which might completely clear HIV from the body.

Compound Q is sometimes referred to as GLQ223 or trichosanthin.

Current use

Compound Q is an unlicensed, experimental drug. Clinical trials are continuing in the USA.

A trial on compound Q was completed at Project Inform, a community based organisation in California. One major problem was that the drug had to be taken by injection. Project Inform was able to recruit a large number of participants and of these six patients developed dementia and one patient died as a result of this therapy. A simultaneous hospital study did not reveal such serious side-effects, but this may have been because of the smaller sample in the group. Neither study found any efficacy, however.

Another study in 56 people with CD4 counts between 100 and 350 reported significant increases in total T-cell counts. A smaller trial reported some evidence of antiviral effects such as temporary reductions in p24 antigen levels and an increase in CD4 percentage.

People treated with compound Q have commonly reported side-effects including headaches, muscle aches, rashes and flu-like symptoms. A few people have experienced a form of allergic reaction during or soon after infusion which may be fatal unless promptly treated. Mental changes ranging from confusion to dementia have also been reported. These may be due to compound Q killing HIV-infected brain cells, or due to killed HIV-infected cells releasing toxins that affect other uninfected brain cells.

Genelabs, who manufacture compound Q, consider that it is too risky for people with CD4 counts below 100. No further developments with this drug have been reported since 1994.

Use in combination

A 148 person study in which people who had taken AZT for at least nine months either remained on AZT, switched to compound Q or started taking it in combination with AZT found no difference between the study arms on an intention-to-treat analysis.

A small study suggested that for people who had taken AZT or ddI for at least one year but were experienced disease progression, adding compound Q to their existing treatment stopped or reversed the decline in their CD4 cell counts over a 6 month period.

Taking it

Compound Q is produced as a liquid which is administered by intravenous infusion. Close medical supervision during infusions is essential as potentially fatal allergic reactions have been seen. Compound Q is not available in Britain.

Current developments

A non-toxic protein derived from trichosanthin, called TAP 29, has been reported to have anti-HIV activity at lower and therefore safer doses than compound Q.

Key research

McGrath reported that GLQ223 exerts a selective inhibitory effect that initially suppresses viral replication and ultimately selectively kills HIV-infected monocytes/macrophages. The exact mechanism of action is unknown.

Ferrari reported that GLQ223 inhibited syncytium formation between infected H9 cells and uninfected Sup-T1 cells from 0.5 to 4 µg/ml and also inhibited HIV replication in H9 and CEM-SS cells at 1 µg/ml, but was toxic for MT-4 cells (HTLV-I-positive), at doses greater than 0.25 µg/ml.

Pulliam investigated the in vitro effect of trichosanthin on human brain cells and macrophages. Brain cells exposed to HIV-infected macrophages or trichosanthin-treated infected macrophages underwent morphological alterations, with the most severe damage observed in brain cells exposed to trichosanthin-treated HIV-infected macrophages. Brain cells exposed to trichosanthin alone, uninfected macrophages or trichosanthin-treated uninfected macrophages were not damaged. These results suggest that neurological complications observed in some HIV-infected people treated with trichosanthin may be due to the release or production of toxic factors from trichosanthin-killed HIV-infected macrophages.

Kahn enrolled 18 (10 AIDS/8 ARC) subjects with CD4 cell count below 350 in a phase I safety and pharmacokinetics study of a single infusion of GLQ223 1, 8, 16, 24, or 36 µg/kg intravenously. 14/18 subjects had prior treatment with AZT. At completion of the trial, no significant changes in immunological or virological parameters were noted. Toxicities included headaches (12/18), muscle aches (10/18) and flu-like symptoms (10/18); these symptoms were controlled within 24-72 hours by the use of paracetamol and/or non-steroidal anti-inflammatory drugs (NSAIDs). 1 patient receiving GLQ223 36 µg/kg developed a neurological reaction similar to acute disseminated encephalomyelitis. Although the patient was intubated and treated with systemic steroids, recovery was slow and incomplete.

Gorelick (1992), Lifson and Kunz reported on a trial in which 56 people with AIDS or ARC (CD4 count 100-350) received 4 weekly infusions of 8, 16, 24, 36, or 50 µg/kg of GLQ223. Protocol amendments allowed patients to continue therapy and to undergo individual dose escalation up to 200 µg/kg. Some participants were treated for over a year and received as many as 34 infusions. Pharmacokinetic evaluation showed target blood levels and duration of exposure were achieved only in those receiving 36 or 50 µg/kg. Statistically significant increases in total T cell count and beta₂-microglobulin levels were observed. 9 participants discontinued due to an adverse event (including one anaphylactoid reaction), most during the first 4 infusions. Almost all patients experienced a transient flu-like syndrome characterized by fever, headache, myalgias and pharyngitis. Allergic reactions characterized primarily by rash, pruritus and rare hypotension were observed, but patients were generally able to continue to receive GLQ223. Frequency of development of antibodies to GLQ223 as well as time to development of antibodies was related to the dose administered. Overall, 63% developed antibodies, with 90% of patients receiving 50 µg/kg becoming antibody positive within 4 weeks. Development of IgG antibodies to GLQ223 appeared to be temporally associated with reduction in side-effects, without evidence of new toxicities. In contrast, changes in levels of selected surrogate markers (e.g. increased CD4 cell count) continued to be observed in the presence of anti-GLQ223 antibodies.

Byers (1990) evaluated 51 people with CD4 cell count below 400 who received 1-3 infusions of a trichosanthin preparation in varying doses (10-30 µg/kg), routes of administration (intramuscularly or intravenously) and dosing intervals. While the investigators reported significant declines in p24 antigenaemia in 16/18 patients who entered with detectable levels, an accompanying editorial cautioned that pre-treatment with steroids may have confounded these measurements, although the researchers point out that only 5/51 patients received steroids. Overall, at completion of the study there was no significant change in CD4 cell count. Side-effects included mental status changes (dementia) in 6 subjects. 2/6 subjects became comatose and 1 died (aspiration pneumonia). 41/51 experienced myalgias, 38/51 had fevers and 23/51 developed rashes more than 36 hours after treatment.

Mayer treated 20 people with AIDS, ARC or asymptomatic HIV infection with 20 µg/kg trichosanthin once every 4 weeks for up to 12 weeks. With the concurrent administration of prostaglandin inhibitors the drug was moderately well tolerated, with most participants experiencing mild arthralgia, hives and malaise. Additionally, 4 participants experienced neurological complications which resolved spontaneously without intervention. 4/20 participants showed progressive although transient reductions in p24 antigen level. 10 people with asymptomatic infection or ARC had significant increases in CD4 percentage and improved delayed hypersensitivity reactions. The group as a whole had a weight gain of 3.2 kg. Improvements in symptoms such as PGL and candidiasis were seen in some participants.

Gorelick (1994) randomized 148 people with at least 9 months prior AZT either to receive continued AZT, to switch to trichosanthin (16 doses, one every 3 weeks, starting at 36, 50 and 100 µg/kg for two weeks each then 150 µg/kg for 10 weeks), or to take trichosanthin in combination with AZT. On intention-to-treat analysis there was no significant difference between the arms in terms of reduction of HIV RNA by quantitative PCR. Combination recipients were significantly more likely to experience a 100-cell increase in CD4 count (67%, versus 38% in the AZT alone group and 44% in the trichosanthin alone group). Two-thirds of trichosanthin recipients received concomitant steroids to reduce allergic responses. Individuals who switched to trichosanthin and did not receive steroids had a significantly longer time to treatment failure than those who continued on AZT monotherapy.

Byers (1991b) conducted a phase II study in which 112 patients (CD4 below 500) who had received AZT or ddI for at least one year and who were failing these drugs had trichosanthin added to their treatment regimen. Participants were infused with trichosanthin (1.2 mg) weekly for 2 weeks then monthly. 65 patients had sufficient CD4 values before (311 +/- 12 days) and after (170 +/- 7 days) trichosanthin to establish changes. On anti-viral drugs alone, patients were losing CD4 cells (1.47 cells/week). After trichosanthin therapy there was a gain of CD4 cells (1.61 cells/week). Serum p24 antigen levels were elevated in 22 patients and these were not influenced by trichosanthin treatment.

Lee-Huang and researchers from the National Cancer Institute have reported the development of a protein, TAP 29, derived from Trichosanthes kirilowii, which displayed in vitro anti-HIV activity as measured by syncytium formation, p24 expression and HIV reverse transcriptase activity, with a therapeutic index over two orders of magnitude higher than that of GLQ223.

References

Byers VS et al. A phase I/II study of trichosanthin treatment of HIV disease. AIDS 4:1189-1196, 1990.

Byers VS et al. Trichosanthin treatment of HIV disease (letter). AIDS 5(9):1150-1151, 1991a.

Byers VS et al. A phase II study of the effect of trichosanthin treatment in combination with zidovudine of HIV disease. 7th Intl Conf AIDS, Florence, abstract WB 2171, 1991b.

Ferrari P et al. Toxicity and activity of purified trichosanthin. AIDS 5(7):865-870, 1991.

Gorelick KJ et al. Phase IB study of GLQ223 in AIDS: surrogate markers. 8th Intl Conf AIDS, Amsterdam, abstract PoB 3442, 1992.

Gorelick K et al. Results of a randomized study of GLQ223 in AIDS and ARC. 10th Intl Conf AIDS, Yokohama, abstract 006B, 1994.

Kahn JO et al. The safety and pharmacokinetics of GLQ223 in subjects with AIDS and AIDS-related complex: a phase I study. AIDS 4:1197-1204, 1990.

Kunz AY et al. Safety of GLQ223 in patients with AIDS and ARC. 8th Intl Conf AIDS, Amsterdam, abstract PuB 7290, 1992.

Lee-Huang S et al. TAP 29: an anti-human immunodeficiency virus protein from Trichosanthes kirilowii that is non-toxic to intact cells. PNAS 88(15):6570-6574, 1991.

Lifson J et al. Anti-GLQ223 antibodies in treated patients with AIDS and ARC. 8th Intl Conf AIDS, Amsterdam, abstract PuB 7314, 1992.

Mayer RA et al. Trichosanthin treatment of HIV-induced immune dysregulation. Eur J Clin Invest 22(2):113-122, 1992.

McGrath MS et al. GLQ223: an inhibitor of human immunodeficiency virus replication in acutely and chronically infected cells of lymphocyte and mononuclear phagocyte lineage. PNAS 86:2844-2848, 1989.

Pinching AJ. Editorial comment: Early trials of GLQ223/trichosanthin: what do they show? AIDS 4:1289-1291, 1990.

Pulliam L et al. Purified trichosanthin (GLQ223) exacerbation of indirect HIV-associated neurotoxicity in vitro. AIDS 5:1237-1232, 1991.